Cutaneous melanoma presents a significant health problem in Western Europe, Australia, and the United States. Americans in the United States currently have approximately a 1 in 75 lifetime risk developing this disease. If melanoma is not excised before it metastasizes the prognosis is poor. Recent evaluation of adjuvant treatments such as interferon alpha indicate the ability to prolong survival in patients with occult metastases, so identification of these patients is highly important. The goal of this proposal is to develop DNA copy number and gene expression markers in the primary tumor that will improve the ability to determine if clinically significant metastasis has occurred by the time of diagnosis. Sentinel lymph node (SLN) biopsy, coupled with several other characteristics of the primary tumor such as thickness and ulceration, are currently the best prognostic indicators. The finding of malignant cells in a sentinel node is a definite sign that the disease has begun to spread. but negative nodes do not guarantee a good outcome. A substantial proportion of all cases of metastatic melanoma come from patients who were SNL- at diagnosis. Thus improving the ability to determine which patients were at elevated risk of progression would have substantial medical benefit. In this project we w ill study primary tumors and corresponding metastases from a cohort of 700 patients that underwent SLN biopsy at UCSF, and an additional set of 100 primary tumors of Acral Lentiginous Melanoma (ALM), to detect DNA copy number changes and gene expression changes that distinguish primary tumors that metastasize from those that do not. Screening for DNA copy aberrations will be performed using array comparative genomic hybridization (array CGH) which has recently been developed in our laboratories. Candidate genes will be identified based on the copy number aberrations, and the prognostic power of the expression levels of these candidates will be evaluated. Using an independent set of tumors from the SLN cohort, we will validate the ability of these markers to: 1) distinguish patients that develop metastases from those that do not, and 2) to distinguish patients with negative sentinel nodes that develop metastases from those with negative nodes that remain disease free.